The long-term objectives of this application are to elucidate the pathological alterations which permit musculoskeletal sepsis to occur and utilize the knowledge of the pathological mechanisms to improve the treatment of musculoskeletal sepsis. This application specifically addresses two types of musculoskeletal sepsis: acute hematogenous osteomyelitis (A.H.O.) and the infected total hip arthroplasty. There are five specific aims: 1) A.H.O. is caused by an alteration of the host defense mechanism; 2) A.H.O. occurs when the oxygen tension in the epiphyseal-metaphyseal junction is reduced; 3) reduction of the oxygen tension inhibits leukocyte interaction with tissue based macrophages; 4) concomittant therapy of antimicrobials with substances that enhance cell membrane fluidity such as pentoxifylline which locally increases chemotaxis and phagocytosis can be currative of A.H.O.; and 5) a delayed, two-staged reconstruction of the infected total hip arthroplasty with an ingrowth arthroplasty is superior and safer than a one-staged reconstruction with antibiotic impregnated bone cement. Aims one and two will be accomplished using the rabbit model of A.H.O. described by Morrissy using light microscopy and transmission electron microscopy, as well as, direct in vivo oxygen tension measurements. It is hypothesized that the absence of peripheral polymorphonuclear leukocytes in the area of the growth plate leaves the responsibility for bacterial phagocytosis and killing of bacteria transported to this anatomic location during a bacteremia to the tissue-based macrophages. Trauma further reduces the very low oxygen tension. Preliminary ultrastructural studies document growth of bacteria in the traumatized zone of provisional calcification. The third and fourth aims will use the rat model of A.H.O. to permit the use of available monoclonal antibodies to define the leukocyte phenotypes and the interaction of lymphocytes and monocytes with tissue based macrophages as well as perivascular cells. The final aim will be addressed in a canine model of a gram-negative bacillary infection about a cemented total hip arthroplasty. It is hypothesized that a one-stage procedure fails to remove all of the glycocalyx and bacteria from the bony femur and acetabulum accounting for a reported failure rate of up to 50%. In contrast, a two-stage procedure with removal of all foreign material permits the host-defense mechanism to eradicate any remaining bacteria following excision of the arthroplasty.